
from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
from plotly.graph_objs import Scatter, Figure, Layout
from plotly.subplots import make_subplots
import plotly
import plotly.express as px
import plotly.graph_objects as go


import sys
import re
import nltk
import time
from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize
from nltk.tokenize import sent_tokenize
from nltk.stem import WordNetLemmatizer

from sklearn.base import BaseEstimator, TransformerMixin
from sklearn.pipeline import Pipeline

from sklearn.feature_extraction.text import TfidfVectorizer, CountVectorizer, TfidfTransformer


nltk.download('stopwords')
nltk.download('wordnet')
nltk.download('punkt')

[nltk_data] Downloading package stopwords to
[nltk_data]     /Users/minyan/nltk_data...
[nltk_data]   Package stopwords is already up-to-date!
[nltk_data] Downloading package wordnet to /Users/minyan/nltk_data...
[nltk_data]   Package wordnet is already up-to-date!
[nltk_data] Downloading package punkt to /Users/minyan/nltk_data...
[nltk_data]   Package punkt is already up-to-date!

True


import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import project_tests as t
import seaborn as sns

%matplotlib inline

df = pd.read_csv('data/user-item-interactions.csv')
df_content = pd.read_csv('data/articles_community.csv')
del df['Unnamed: 0']
del df_content['Unnamed: 0']

# Show df to get an idea of the data
df.head()


# Show df_content to get an idea of the data
df_content.head()


#dataframe of number of articles each user interacts with. Removeds duplicate
print(df.groupby('email')['article_id'].nunique())
#dataframe of number of articles each user interacts with. Duplicate articles included
print(df.groupby('email')['article_id'].count())

email
0000b6387a0366322d7fbfc6434af145adf7fed1    12
001055fc0bb67f71e8fa17002342b256a30254cd     4
00148e4911c7e04eeff8def7bbbdaf1c59c2c621     3
001a852ecbd6cc12ab77a785efa137b2646505fe     5
001fc95b90da5c3cb12c501d201a915e4f093290     2
                                            ..
ffc6cfa435937ca0df967b44e9178439d04e3537     1
ffc96f8fbb35aac4cb0029332b0fc78e7766bb5d     2
ffe3d0543c9046d35c2ee3724ea9d774dff98a32    27
fff9fc3ec67bd18ed57a34ed1e67410942c4cd81     7
fffb93a166547448a0ff0232558118d59395fecd     6
Name: article_id, Length: 5148, dtype: int64
email
0000b6387a0366322d7fbfc6434af145adf7fed1    13
001055fc0bb67f71e8fa17002342b256a30254cd     4
00148e4911c7e04eeff8def7bbbdaf1c59c2c621     3
001a852ecbd6cc12ab77a785efa137b2646505fe     6
001fc95b90da5c3cb12c501d201a915e4f093290     2
                                            ..
ffc6cfa435937ca0df967b44e9178439d04e3537     2
ffc96f8fbb35aac4cb0029332b0fc78e7766bb5d     4
ffe3d0543c9046d35c2ee3724ea9d774dff98a32    32
fff9fc3ec67bd18ed57a34ed1e67410942c4cd81    10
fffb93a166547448a0ff0232558118d59395fecd    13
Name: article_id, Length: 5148, dtype: int64


#build the dataframe records the number of articles each user interacts with. Duplicate article counts have been removed
num_articles_by_user = df.groupby('email')['article_id'].nunique()


#descriptive statistics of dataset overview
num_articles_by_user.describe()

count    5148.000000
mean        6.540210
std         9.990676
min         1.000000
25%         1.000000
50%         3.000000
75%         7.000000
max       135.000000
Name: article_id, dtype: float64


#build the dataframe records the number of articles each user interacts with. Duplicate article counts have been removed
num_articles_by_any_user = df.groupby('email')['email'].count()


#descriptive statistics of dataset overview
num_articles_by_any_user.describe()

count    5148.000000
mean        8.930847
std        16.802267
min         1.000000
25%         1.000000
50%         3.000000
75%         9.000000
max       364.000000
Name: email, dtype: float64


print(num_articles_by_user.index)
print(num_articles_by_user.values)

Index(['0000b6387a0366322d7fbfc6434af145adf7fed1',
       '001055fc0bb67f71e8fa17002342b256a30254cd',
       '00148e4911c7e04eeff8def7bbbdaf1c59c2c621',
       '001a852ecbd6cc12ab77a785efa137b2646505fe',
       '001fc95b90da5c3cb12c501d201a915e4f093290',
       '0042719415c4fca7d30bd2d4e9d17c5fc570de13',
       '00772abe2d0b269b2336fc27f0f4d7cb1d2b65d7',
       '008ba1d5b4ebf54babf516a2d5aa43e184865da5',
       '008ca24b82c41d513b3799d09ae276d37f92ce72',
       '008dfc7a327b5186244caec48e0ab61610a0c660',
       ...
       'ff6e82c763fe2443643e48a03e239eb635f406dc',
       'ff7a0f59ba022102ad22981141a7182c4d8273c3',
       'ff833869969184d86f870f98405e7988eccc2309',
       'ff979e07f9d906a32ba35a9b75fd9585f6306dbc',
       'ffaefa3a1bc2d074d9a14c9924d4e67a46c35410',
       'ffc6cfa435937ca0df967b44e9178439d04e3537',
       'ffc96f8fbb35aac4cb0029332b0fc78e7766bb5d',
       'ffe3d0543c9046d35c2ee3724ea9d774dff98a32',
       'fff9fc3ec67bd18ed57a34ed1e67410942c4cd81',
       'fffb93a166547448a0ff0232558118d59395fecd'],
      dtype='object', name='email', length=5148)
[12  4  3 ... 27  7  6]


df_num_articles_by_users = pd.DataFrame({'user_id':num_articles_by_user.index, 'number_of_articles_by_user':num_articles_by_user.values})


fig = px.histogram(df_num_articles_by_users,x='number_of_articles_by_user')
fig.show()


# Fill in the median and maximum number of user_article interactios below

median_val = num_articles_by_user.median()# 50% of individuals interact with ____ number of articles or fewer.
max_views_by_user =num_articles_by_any_user.max() # The maximum number of user-article interactions by any 1 user is ______.
print(median_val)
print(max_views_by_user)

3.0
364


# Find and explore duplicate articles
print(df_content.shape)
print(df_content['article_id'].nunique())

(1056, 5)
1051


# Find and explore duplicate articles, identify article_id with duplicates
article_content=df_content.groupby('article_id')['article_id'].count()
article_content[article_content.values>1]

article_id
50     2
221    2
232    2
398    2
577    2
Name: article_id, dtype: int64


# Remove any rows that have the same article_id - only keep the first
df_content = df_content.drop_duplicates(subset='article_id', keep='first')
print(df_content.shape)

(1051, 5)


#The number of unique articles that have at least one interaction
df_article = df.groupby('article_id')['email'].count()
unique_a = df_article[df_article.values>0].count()
print(unique_a)

714


# The number of unique articles on the IBM platform
total_a = df_content['article_id'].nunique()
print(total_a)

1051


# The number of unique users
unique_u = df['email'].dropna()
unique_u = unique_u.nunique()
print(unique_u)

5148


# The number of user-article interactions
user_article_i = len(df)
print(user_article_i)

45993


unique_articles = unique_a# The number of unique articles that have at least one interaction
total_articles = total_a# The number of unique articles on the IBM platform
unique_users = unique_u # The number of unique users
user_article_interactions = user_article_i# The number of user-article interactions


# The most viewed article in the dataset as a string with one value following the decimal 
df.groupby('article_id')['article_id'].count().sort_values(ascending=False)

article_id
1429.0    937
1330.0    927
1431.0    671
1427.0    643
1364.0    627
         ... 
1113.0      1
1119.0      1
984.0       1
1127.0      1
1266.0      1
Name: article_id, Length: 714, dtype: int64


print(df.groupby('article_id')['article_id'].count().sort_values(ascending=False).index[0])
print(df.groupby('article_id')['article_id'].count().sort_values(ascending=False).values[0])

1429.0
937


most_viewed_article_id = df.groupby('article_id')['article_id'].count().sort_values(ascending=False).index[0]# The most viewed article in the dataset as a string with one value following the decimal 
most_viewed_article_id = str(most_viewed_article_id)
max_views = df.groupby('article_id')['article_id'].count().sort_values(ascending=False).values[0]# The most viewed article in the dataset was viewed how many times?


## No need to change the code here - this will be helpful for later parts of the notebook
# Run this cell to map the user email to a user_id column and remove the email column

def email_mapper():
    coded_dict = dict()
    cter = 1
    email_encoded = []
    
    for val in df['email']:
        if val not in coded_dict:
            coded_dict[val] = cter
            cter+=1
        
        email_encoded.append(coded_dict[val])
    return email_encoded

email_encoded = email_mapper()
del df['email']
df['user_id'] = email_encoded

# show header
df.head()


## If you stored all your results in the variable names above, 
## you shouldn't need to change anything in this cell

sol_1_dict = {
    '`50% of individuals have _____ or fewer interactions.`': median_val,
    '`The total number of user-article interactions in the dataset is ______.`': user_article_interactions,
    '`The maximum number of user-article interactions by any 1 user is ______.`': max_views_by_user,
    '`The most viewed article in the dataset was viewed _____ times.`': max_views,
    '`The article_id of the most viewed article is ______.`': most_viewed_article_id,
    '`The number of unique articles that have at least 1 rating ______.`': unique_articles,
    '`The number of unique users in the dataset is ______`': unique_users,
    '`The number of unique articles on the IBM platform`': total_articles
}

# Test your dictionary against the solution
t.sol_1_test(sol_1_dict)

It looks like you have everything right here! Nice job!


df.groupby('article_id')['article_id'].count().sort_values(ascending=False).index

Float64Index([1429.0, 1330.0, 1431.0, 1427.0, 1364.0, 1314.0, 1293.0, 1170.0,
              1162.0, 1304.0,
              ...
               662.0, 1200.0, 1092.0,  653.0, 1344.0, 1113.0, 1119.0,  984.0,
              1127.0, 1266.0],
             dtype='float64', name='article_id', length=714)


ranked_interactions = df.groupby('article_id')['article_id'].count().sort_values(ascending=False).index
number_interactions = df.groupby('article_id')['article_id'].count().sort_values(ascending=False).values
#pull article title and article_id for each article
article_titles = df[['article_id','title']].drop_duplicates()
#build ranked articles 
ranked_articles_df = pd.DataFrame({'ranked_article_id': ranked_interactions, 'num_ratings': number_interactions})
ranked_articles_df = pd.merge(ranked_articles_df, article_titles, left_on='ranked_article_id', right_on='article_id').drop('article_id', axis=1)


ranked_articles_df.head()


def get_top_articles(n, df=df):
    '''
    INPUT:
    n - (int) the number of top articles to return
    df - (pandas dataframe) df as defined at the top of the notebook 
    
    OUTPUT:
    top_articles - (list) A list of the top 'n' article titles 
    
    '''
    # Your code here
    #pull the ranking, and number of interactions for each articles
    top_articles = list(df.groupby('title').count().sort_values(by='user_id',ascending=False).head(n).index)
   
    return top_articles # Return the top article titles from df (not df_content)

def get_top_article_ids(n, df=df):
    '''
    INPUT:
    n - (int) the number of top articles to return
    df - (pandas dataframe) df as defined at the top of the notebook 
    
    OUTPUT:
    top_articles - (list) A list of the top 'n' article titles 
    
    '''
    # Your code here
    #pull the ranking, and number of interactions for each articles
    top_articles = list(df.groupby('article_id').count().sort_values(by='user_id',ascending=False).head(n).index)
   
    return top_articles # Return the top article ids


print(get_top_articles(10))
print(get_top_article_ids(10))

['use deep learning for image classification', 'insights from new york car accident reports', 'visualize car data with brunel', 'use xgboost, scikit-learn & ibm watson machine learning apis', 'predicting churn with the spss random tree algorithm', 'healthcare python streaming application demo', 'finding optimal locations of new store using decision optimization', 'apache spark lab, part 1: basic concepts', 'analyze energy consumption in buildings', 'gosales transactions for logistic regression model']
[1429.0, 1330.0, 1431.0, 1427.0, 1364.0, 1314.0, 1293.0, 1170.0, 1162.0, 1304.0]


# Test your function by returning the top 5, 10, and 20 articles
top_5 = get_top_articles(5)
top_10 = get_top_articles(10)
top_20 = get_top_articles(20)

# Test each of your three lists from above
t.sol_2_test(get_top_articles)

Your top_5 looks like the solution list! Nice job.
Your top_10 looks like the solution list! Nice job.
Your top_20 looks like the solution list! Nice job.


# create the user-article matrix with 1's and 0's

def create_user_item_matrix(df):
    '''
    INPUT:
    df - pandas dataframe with article_id, title, user_id columns
    
    OUTPUT:
    user_item - user item matrix 
    
    Description:
    Return a matrix with user ids as rows and article ids on the columns with 1 values where a user interacted with 
    an article and a 0 otherwise
    '''
    # Fill in the function here
    # Create user-by-item matrix
    user_by_article = df.groupby(['user_id', 'article_id'])['article_id'].count().unstack()
    user_by_article[user_by_article>1]=1
    user_item=user_by_article.fillna(0)
    return user_item # return the user_item matrix 

user_item = create_user_item_matrix(df)


user_item


## Tests: You should just need to run this cell.  Don't change the code.
assert user_item.shape[0] == 5149, "Oops!  The number of users in the user-article matrix doesn't look right."
assert user_item.shape[1] == 714, "Oops!  The number of articles in the user-article matrix doesn't look right."
assert user_item.sum(axis=1)[1] == 36, "Oops!  The number of articles seen by user 1 doesn't look right."
print("You have passed our quick tests!  Please proceed!")

You have passed our quick tests!  Please proceed!


def find_similar_users(user_id, user_item=user_item):
    '''
    INPUT:
    user_id - (int) a user_id
    user_item - (pandas dataframe) matrix of users by articles: 
                1's when a user has interacted with an article, 0 otherwise
    
    OUTPUT:
    similar_users - (list) an ordered list where the closest users (largest dot product users)
                    are listed first
    
    Description:
    Computes the similarity of every pair of users based on the dot product
    Returns an ordered
    
    '''
    user_id_series=np.array(user_item.index)
    # compute similarity of each user to the provided user
    val=user_item[user_item.index==user_id]
    val_similarity = val.dot(np.transpose(user_item))
    # sort by similarity
    # function: argsort, which returns the indices of the sorted elements
    val_similarity_indices = np.argsort(-val_similarity)
    # create list of just the ids
    val_similarity_ids = list(user_id_series[val_similarity_indices])
    # remove the own user's id
    most_similar_users=[]
    for i in range(len(user_item)):
        if val_similarity_ids[0][i] != user_id:
             most_similar_users.append(val_similarity_ids[0][i])
       
    return most_similar_users # return a list of the users in order from most to least similar


# Do a spot check of your function
print("The 10 most similar users to user 1 are: {}".format(find_similar_users(1)[:10]))
print("The 5 most similar users to user 3933 are: {}".format(find_similar_users(3933)[:5]))
print("The 3 most similar users to user 46 are: {}".format(find_similar_users(46)[:3]))

The 10 most similar users to user 1 are: [3933, 3782, 23, 203, 4459, 3870, 131, 46, 4201, 49]
The 5 most similar users to user 3933 are: [1, 23, 3782, 4459, 203]
The 3 most similar users to user 46 are: [4201, 3782, 23]


def get_article_names(article_ids, df=df):
    '''
    INPUT:
    article_ids - (list) a list of article ids
    df - (pandas dataframe) df as defined at the top of the notebook
    
    OUTPUT:
    article_names - (list) a list of article names associated with the list of article ids 
                    (this is identified by the title column)
    '''
    
     # pull the movie titles based on the indices
    article_names = set(df[df['article_id'].isin(article_ids)]['title'])
    return article_names # Return the article names associated with list of article ids


def get_user_articles(user_id, user_item=user_item):
    '''
    INPUT:
    user_id - (int) a user id
    user_item - (pandas dataframe) matrix of users by articles: 
                1's when a user has interacted with an article, 0 otherwise
    
    OUTPUT:
    article_ids - (list) a list of the article ids seen by the user
    article_names - (list) a list of article names associated with the list of article ids 
    
    Description:
    Provides a list of the article_ids and article titles that have been seen by a user
    '''
    # Your code here
    # Pull only the reviews the user has seen
    temp = user_item.loc[user_id][user_item.loc[user_id].values== 1].index
    article_ids = np.array(temp)
    article_names = get_article_names(article_ids, df=df)
    return article_ids, article_names # return the ids and names


def user_user_recs(user_id, m=10):
    '''
    INPUT:
    user_id - (int) a user id
    m - (int) the number of recommendations you want for the user
    
    OUTPUT:
    recs - (list) a list of recommendations for the user
    
    Description:
    Loops through the users based on closeness to the input user_id
    For each user - finds articles the user hasn't seen before and provides them as recs
    Does this until m recommendations are found
    
    Notes:
    Users who are the same closeness are chosen arbitrarily as the 'next' user
    
    For the user where the number of recommended articles starts below m 
    and ends exceeding m, the last items are chosen arbitrarily
    
    '''
    
    # Keep the recommended articles ids here
    recs_ids = np.array([])
    #recs = np.array([])
    
    # Your code here
    most_similar_users = find_similar_users(user_id, user_item=user_item)
    # articles been read by target users
    target_user_article_ids, target_user_article_names= get_user_articles(user_id, user_item=user_item)
    
   
    
    # Go through the neighbors and identify movies they like the user hasn't seen
    for neighbor in most_similar_users:
        neighbs_likes_ids,neighbs_likes_names = get_user_articles(neighbor,user_item=user_item)
        #Obtain recommendations for each neighbor
        # These will be the recommendations - continue until 10 recs 
        new_recs_ids = np.setdiff1d(neighbs_likes_ids, target_user_article_ids, assume_unique=True)
        #for article in new_recs_ids:
            # Update recs with new recs
        recs_ids = np.unique(np.concatenate([recs_ids, new_recs_ids], axis=0))
        
            # If we have enough recommendations exit the loop
        if len(recs_ids) > m-1:
            break
    
    # Pull movie titles using movie ids
    recs = recs_ids[:m]
    
    return recs # return your recommendations for this user_id


####alternative to above,it returns 10!!!!!
####################################!!!!!!#####################################

def get_article_names(article_ids, df=df):
    '''
    INPUT:
    article_ids - (list) a list of article ids
    df - (pandas dataframe) df as defined at the top of the notebook
    
    OUTPUT:
    article_names - (list) a list of article names associated with the list of article ids 
                    (this is identified by the title column)
    '''
    
     # pull the movie titles based on the indices
    article_names = set(df[df['article_id'].isin(article_ids)]['title'])
    return article_names # Return the article names associated with list of article ids


def get_user_articles(user_id, user_item=user_item):
    '''
    INPUT:
    user_id - (int) a user id
    user_item - (pandas dataframe) matrix of users by articles: 
                1's when a user has interacted with an article, 0 otherwise
    
    OUTPUT:
    article_ids - (list) a list of the article ids seen by the user
    article_names - (list) a list of article names associated with the list of article ids 
    
    Description:
    Provides a list of the article_ids and article titles that have been seen by a user
    '''
    # Your code here
    # Pull only the reviews the user has seen
    temp = user_item.loc[user_id][user_item.loc[user_id].values== 1].index
    article_ids = np.array(temp)
    article_names = get_article_names(article_ids, df=df)
    return article_ids, article_names # return the ids and names


def user_user_recs(user_id, m=10):
    '''
    INPUT:
    user_id - (int) a user id
    m - (int) the number of recommendations you want for the user
    
    OUTPUT:
    recs - (list) a list of recommendations for the user
    
    Description:
    Loops through the users based on closeness to the input user_id
    For each user - finds articles the user hasn't seen before and provides them as recs
    Does this until m recommendations are found
    
    Notes:
    Users who are the same closeness are chosen arbitrarily as the 'next' user
    
    For the user where the number of recommended articles starts below m 
    and ends exceeding m, the last items are chosen arbitrarily
    
    '''
    
    # Keep the recommended articles ids here
    #recs_ids = np.array([])
    recs_ids = []
    #recs = np.array([])
    
    # Your code here
    most_similar_users = find_similar_users(user_id, user_item=user_item)
    # articles been read by target users
    target_user_article_ids, target_user_article_names= get_user_articles(user_id, user_item=user_item)
    
    # Go through the neighbors and identify movies they like the user hasn't seen
    for neighbor in most_similar_users:
        neighbs_likes_ids,neighbs_likes_names = get_user_articles(neighbor,user_item=user_item)
        #Obtain recommendations for each neighbor
        # These will be the recommendations - continue until 10 recs 
        new_recs_ids = np.setdiff1d(neighbs_likes_ids, target_user_article_ids, assume_unique=True)
        new_recs_ids = list(new_recs_ids)
        for article in new_recs_ids:
            if article not in recs_ids:
                recs_ids.append(article)
                if len(recs_ids) > m-1:
                    break
        if len(recs_ids) > m-1:
                    break
    # Pull movie titles using movie ids
    recs = recs_ids
    return recs # return your recommendations for this user_id


# Check Results
get_article_names(user_user_recs(1, 10)) # Return 10 recommendations for user 1

{'deep forest: towards an alternative to deep neural networks',
 'experience iot with coursera',
 'got zip code data? prep it for analytics. – ibm watson data lab – medium',
 'graph-based machine learning',
 'higher-order logistic regression for large datasets',
 'the 3 kinds of context: machine learning and the art of the frame',
 'this week in data science (april 18, 2017)',
 'timeseries data analysis of iot events by using jupyter notebook',
 'using brunel in ipython/jupyter notebooks',
 'using machine learning to predict parking difficulty'}


get_user_articles(20)[0]

array([ 232.,  844., 1320.])


get_user_articles(2)[1]

{'build a python app on the streaming analytics service',
 'gosales transactions for naive bayes model',
 'healthcare python streaming application demo',
 'use r dataframes & ibm watson natural language understanding',
 'use xgboost, scikit-learn & ibm watson machine learning apis',
 'using deep learning to reconstruct high-resolution audio'}


# Test your functions here - No need to change this code - just run this cell
assert set(get_article_names([1024.0, 1176.0, 1305.0, 1314.0, 1422.0, 1427.0])) == set(['using deep learning to reconstruct high-resolution audio', 'build a python app on the streaming analytics service', 'gosales transactions for naive bayes model', 'healthcare python streaming application demo', 'use r dataframes & ibm watson natural language understanding', 'use xgboost, scikit-learn & ibm watson machine learning apis']), "Oops! Your the get_article_names function doesn't work quite how we expect."
assert set(get_article_names([1320.0, 232.0, 844.0])) == set(['housing (2015): united states demographic measures','self-service data preparation with ibm data refinery','use the cloudant-spark connector in python notebook']), "Oops! Your the get_article_names function doesn't work quite how we expect."
assert set(get_user_articles(20)[0]) == set([232.,  844., 1320.])
assert set(get_user_articles(20)[1]) == set(['housing (2015): united states demographic measures', 'self-service data preparation with ibm data refinery','use the cloudant-spark connector in python notebook'])
assert set(get_user_articles(2)[0]) == set([1024., 1176., 1305., 1314., 1422., 1427.])
assert set(get_user_articles(2)[1]) == set(['build a python app on the streaming analytics service','gosales transactions for naive bayes model','healthcare python streaming application demo','use r dataframes & ibm watson natural language understanding','use xgboost, scikit-learn & ibm watson machine learning apis','using deep learning to reconstruct high-resolution audio'])
print("If this is all you see, you passed all of our tests!  Nice job!")

If this is all you see, you passed all of our tests!  Nice job!


#######alternatives##########
#use get_top_article_ids function for article orders 
#############################


def get_top_sorted_users(user_id, df=df, user_item=user_item):
    '''
    INPUT:
    user_id - (int)
    df - (pandas dataframe) df as defined at the top of the notebook 
    user_item - (pandas dataframe) matrix of users by articles: 
            1's when a user has interacted with an article, 0 otherwise
    
            
    OUTPUT:
    neighbors_df - (pandas dataframe) a dataframe with:
                    neighbor_id - is a neighbor user_id
                    similarity - measure of the similarity of each user to the provided user_id
                    num_interactions - the number of articles viewed by the user - if a u
                    
    Other Details - sort the neighbors_df by the similarity and then by number of interactions where 
                    highest of each is higher in the dataframe
     
    '''
    # Your code here
    # compute similarity of each user to the provided user
    val=user_item[user_item.index==user_id]
    val_similarity = val.dot(np.transpose(user_item))
    val_similarity = np.transpose(val_similarity)
    val_s=pd.DataFrame(val_similarity)
    val_s.columns=['similarity']

    # compute interactions for each users
    num_interactions = np.transpose(user_item).sum()

    neighbor_df = pd.DataFrame({'num_interactions': num_interactions})
    neighbor_df = neighbor_df.join(val_s)
    neighbor_df.reset_index(inplace=True)
    neighbor_df = neighbor_df.rename(columns = {'user_id':'neighbor_id'})
    
    # sort by similarity, and number of article interactions
    ranked_df = neighbor_df.sort_values(['similarity', 'num_interactions'], ascending=False)
    
    # remove the own user's id
    neighbors_df = ranked_df[ranked_df.neighbor_id != user_id]
        
    return neighbors_df # Return the dataframe specified in the doc_string


def user_user_recs_part2(user_id, m=10):
    '''
    INPUT:
    user_id - (int) a user id
    m - (int) the number of recommendations you want for the user
    
    OUTPUT:
    recs - (list) a list of recommendations for the user by article id
    rec_names - (list) a list of recommendations for the user by article title
    
    Description:
    Loops through the users based on closeness to the input user_id
    For each user - finds articles the user hasn't seen before and provides them as recs
    Does this until m recommendations are found
    
    Notes:
    * Choose the users that have the most total article interactions 
    before choosing those with fewer article interactions.

    * Choose articles with the articles with the most total interactions 
    before choosing those with fewer total interactions. 
   
    '''
    # Your code here
     # Keep the recommended articles ids here
    recs_ids = []
    #recs = np.array([])
    
    # Your code here
    neighbors_df = get_top_sorted_users(user_id, df=df, user_item=user_item)
    most_similar_users = neighbors_df['neighbor_id']
    
    # articles been read by target users
    target_user_article_ids, target_user_article_names= get_user_articles(user_id, user_item=user_item)
    
    # Go through the neighbors and identify movies they like the user hasn't seen
    for neighbor in most_similar_users:
        neighbs_likes_ids,neighbs_likes_names = get_user_articles(neighbor,user_item=user_item)
        sorted_articles = pd.DataFrame()
        for a in neighbs_likes_ids:
            user_interactions = user_item[:][a].sum()
            sorted_df= pd.DataFrame({'neighbs_likes_ids':[a],'user_interactions':[user_interactions]})
            sorted_articles = sorted_df.append(sorted_df)

        sorted_articles=sorted_articles.sort_values(['user_interactions'],ascending=False)
        sorted_articles_ids=np.array(sorted_articles['neighbs_likes_ids'])
        #Obtain recommendations for each neighbor
        # These will be the recommendations - continue until 10 recs 
        new_recs_ids = np.setdiff1d(sorted_articles_ids, target_user_article_ids, assume_unique=True)
        new_recs_ids = list(new_recs_ids)
        for article in new_recs_ids:
            if article not in recs_ids:
                recs_ids.append(article)
                if len(recs_ids) > m-1:
                    break
        if len(recs_ids) > m-1:
            break       
        
        
    # Pull movie titles using movie ids
    recs = recs_ids
    rec_names = get_article_names(recs, df=df)
    return recs, rec_names


#make recommendations using User-User Based Collaborative Filtering
def all_recommendations(num_recs=10):
    '''
    INPUT 
        num_recs (int) the (max) number of recommendations for each user
    OUTPUT
        all_recs - a dictionary where each key is a user_id and the value is an array of recommended movie titles
    '''
    
    # All the users we need to make recommendations for
    users = np.unique(df['user_id'])
    n_users = len(users)
    
    #Store all recommendations in this dictionary
    all_recs = dict()
    
    # Make the recommendations for each user
    for user in users[0:500]:
        all_recs[user] = user_user_recs_part2(user, num_recs)
    
    return all_recs

all_recs = all_recommendations(10)


all_recs.items()


# Quick spot check - don't change this code - just use it to test your functions
rec_ids, rec_names = user_user_recs_part2(20, 10)
print("The top 10 recommendations for user 20 are the following article ids:")
print(rec_ids)
print()
print("The top 10 recommendations for user 20 are the following article names:")
print(rec_names)

The top 10 recommendations for user 20 are the following article ids:
[1432.0, 1444.0, 1437.0, 1439.0, 1396.0, 1410.0, 1436.0, 1431.0, 1429.0, 1435.0]

The top 10 recommendations for user 20 are the following article names:
{'working with ibm cloud object storage in r', 'what caused the challenger disaster?', 'worldwide fuel oil consumption by household (in 1000 metric tons)', 'visualize car data with brunel', 'uci: sms spam collection', 'watson assistant workspace analysis with user logs', 'times world university ranking analysis', 'use deep learning for image classification', 'welcome to pixiedust', 'visualize data with the matplotlib library'}


#get_top_sorted_users(1, df=df, user_item=user_item).iloc[0]['neighbor_id']
get_top_sorted_users(131, df=df, user_item=user_item).head(30)


### Tests with a dictionary of results

user1_most_sim = get_top_sorted_users(1, df=df, user_item=user_item).iloc[0]['neighbor_id']# Find the user that is most similar to user 1 
user131_10th_sim = get_top_sorted_users(131, df=df, user_item=user_item).iloc[10]['neighbor_id']# Find the 10th most similar user to user 131


## Dictionary Test Here
sol_5_dict = {
    'The user that is most similar to user 1.': user1_most_sim, 
    'The user that is the 10th most similar to user 131': user131_10th_sim,
}

t.sol_5_test(sol_5_dict)

This all looks good!  Nice job!


new_user = '0.0'

# What would your recommendations be for this new user '0.0'?  As a new user, they have no observed articles.
# Provide a list of the top 10 article ids you would give to 
new_user_recs = get_top_article_ids(10, df=df)# Your recommendations here


get_top_article_ids(10, df=df)

[1429.0,
 1330.0,
 1431.0,
 1427.0,
 1364.0,
 1314.0,
 1293.0,
 1170.0,
 1162.0,
 1304.0]


assert set(new_user_recs) == set([1429.0,1330.0,1431.0,1427.0,1364.0,1314.0,1293.0,1170.0,1162.0,1304.0]), "Oops!  It makes sense that in this case we would want to recommend the most popular articles, because we don't know anything about these users."

print("That's right!  Nice job!")

That's right!  Nice job!


#Before finding recommendations, rank the users' article interactions from highest to lowest. You will move through the movies in this order looking for other similar movies.


# create a dataframe, ranked by number of interactions of each articles for each user 
ranked_user_article = df.groupby(['user_id','article_id'])['article_id'].count().sort_values(ascending=False)
df1 = pd.DataFrame(ranked_user_article)
df1.rename(columns={'article_id': 'Counts'}, inplace=True)
df1 = df1.reset_index().sort_values(by=['Counts','user_id'],ascending=[False,True])


def tokenize(text):
    '''
    INPUT:
        text:raw message
    OUTPUT:
        clean_tokens:tokenized words
    DESCRIPTION:
        The function is to process the scentence, normalize texts, tokenize texts.
        Convert all cases to lower cases, remove extra space,stop words, and 
        reduce words to their root form.
    '''
    clean_tokens=[]    
    #remove punctuation,normalize case to lower cases, and remove extra space
    text = re.sub(r"[^a-zA-Z0-9]"," ",text.lower()).strip()
    
    #tokenize text
    tokens=word_tokenize(text)
    
    for w in tokens:  
        #remove stop words
        if w not in stopwords.words("english"):
        #lemmatization
        #reduce words to their root form
            lemmed = WordNetLemmatizer().lemmatize(w)
            clean_tokens.append(lemmed)
    return clean_tokens


####s2
pipeline = Pipeline([
        ('vectorizer', CountVectorizer(tokenizer=tokenize)),
        ('tfidf', TfidfTransformer(smooth_idf=False))
        ])


def find_similar_articles(article_id, df):
    '''
    INPUT
    article_id - a article_id
    movies_df - original movies dataframe
    OUTPUT
    similar_movies - an array of the most similar movies by title
    '''
    #drop duplicate articles and get article titles
    new_df = df.drop_duplicates(subset='article_id')
    titles = np.array(new_df['title'])
    
    #use pipeline defined above to convert texts to numeric matrix
    tfidf = pipeline.fit_transform(titles)
    # convert sparse matrix to numpy array
    titles_tfidf = tfidf.toarray()

    # dot product to get similar articles
    dot_prod_titles = titles_tfidf.dot(np.transpose(titles_tfidf))

    # find the row of each article id
    article_idx = np.where(new_df['article_id'] == article_id)[0][0]

    # find the most similar movie indices - to start I said they need to be the same for all content
    similar_idxs = np.where(dot_prod_titles[article_idx] >= 0.2*np.max(dot_prod_titles[article_idx]))[0]
    # remove article itself
    similar_idxs = np.setdiff1d(similar_idxs,article_idx)
    
    # pull the titles based on the indices
    similar_articles = np.array(new_df.iloc[similar_idxs, ]['title'])

    return similar_articles


 def make_recommendations(_id, _id_type='article', df=df, rec_num=5):
        '''
        INPUT:
        _id - either a user or movie id (int)
        _id_type - "movie" or "user" (str)
        rec_num - number of recommendations to return (int)

        OUTPUT:
        recs - (array) a list or numpy array of recommended movies like the
                       given movie, or recs for a user_id given
        '''
        # if the user is available from the matrix factorization data,
        # I will use this and rank movies based on the predicted values
        # For use with user indexing
        new_df1 = df.drop_duplicates(subset='user_id')
        user_id_series = np.array(new_df1['user_id'])
        
        new_df2 = df.drop_duplicates(subset='article_id')
        article_ids_series = np.array(new_df2['article_id'])
        
        rec_ids, rec_names = None, None
        if _id_type == 'user':
            if _id in user_id_series:
                # Get the index of which row the user is in for use in U matrix
                rec_ids, rec_names = user_user_recs_part2(_id, m=rec_num)
                rec_names = list(rec_names)
            else:
                # if we don't have this user, give just top ratings back
                rec_names = list(get_top_articles(rec_num))
                print("Because this user wasn't in our database, we are giving back the top article recommendations for all users.")

        # Find similar movies if it is a movie that is passed
        else:
            if _id in article_ids_series:
                rec_names = list(find_similar_articles(_id, df))[:rec_num]
            else:
                print("That movie doesn't exist in our database.  Sorry, we don't have any recommendations for you.")

        return rec_names


# make recommendations for a brand new user
rec1=make_recommendations(0, _id_type='user', df=df, rec_num=5)
print('For the new user, here are the recommendations:{}'.format(str(rec1)))


# make a recommendations for a user who only has interacted with article id '1427.0'
rec2=make_recommendations(1427.0, _id_type='article', df=df, rec_num=5)
print('For a user who interacts with the given article , here are the recommendations:{}'.format(str(rec2)))

Because this user wasn't in our database, we are giving back the top article recommendations for all users.
For the new user, here are the recommendations:['use deep learning for image classification', 'insights from new york car accident reports', 'visualize car data with brunel', 'use xgboost, scikit-learn & ibm watson machine learning apis', 'predicting churn with the spss random tree algorithm']
For a user who interacts with the given article , here are the recommendations:['use apache systemml and spark for machine learning', 'use r dataframes & ibm watson natural language understanding', 'python machine learning: scikit-learn tutorial', 'learn tensorflow and deep learning together and now!', 'learn about data science in world of watson']


# Load the matrix here
user_item_matrix = pd.read_pickle('user_item_matrix.p')


# quick look at the matrix
user_item_matrix.shape

(5149, 714)


# Perform SVD on the User-Item Matrix Here

u, s, vt = np.linalg.svd(user_item_matrix, full_matrices=True)# use the built in to get the three matrices


u.shape

(5149, 5149)


# Change the dimensions of u, s, and vt as necessary to use four latent features
# update the shape of u and store in u_new
u_new = u[:,:len(s)]# change the shape of u here

# update the shape of s and store in s_new
s_new =np.diag(s) #change the shape of s as necessary 

# Because we are using 4 latent features and there are only 4 movies, 
# vt and vt_new are the same
vt_new = vt# change the shape of vt as necessary


num_latent_feats = np.arange(10,700+10,20)
sum_errs = []

for k in num_latent_feats:
    # restructure with k latent features
    s_new, u_new, vt_new = np.diag(s[:k]), u[:, :k], vt[:k, :]
    
    # take dot product
    user_item_est = np.around(np.dot(np.dot(u_new, s_new), vt_new))
    
    # compute error for each prediction to actual value
    diffs = np.subtract(user_item_matrix, user_item_est)
    
    # total errors and keep track of them
    err = np.sum(np.sum(np.abs(diffs)))
    sum_errs.append(err)
    
    
plt.plot(num_latent_feats, np.array(sum_errs)/df.shape[0]);
plt.xlabel('Number of Latent Features');
plt.ylabel('Accuracy');
plt.title('Accuracy vs. Number of Latent Features');


df_train = df.head(40000)
df_test = df.tail(5993)

def create_test_and_train_user_item(df_train, df_test):
    '''
    INPUT:
    df_train - training dataframe
    df_test - test dataframe
    
    OUTPUT:
    user_item_train - a user-item matrix of the training dataframe 
                      (unique users for each row and unique articles for each column)
    user_item_test - a user-item matrix of the testing dataframe 
                    (unique users for each row and unique articles for each column)
    test_idx - all of the test user ids
    test_arts - all of the test article ids
    
    '''
    # Your code here
    user_item_train = create_user_item_matrix(df_train)
    user_item_test = create_user_item_matrix(df_test)

    test_idx = user_item_test.index
    test_arts = user_item_test.columns
    
    return user_item_train, user_item_test, test_idx, test_arts

user_item_train, user_item_test, test_idx, test_arts = create_test_and_train_user_item(df_train, df_test)


#cold start probelm
#users in the test set but not in the training set can not be predicted
#articles in the test set but not in the training set can not be predicted


train_idx1 = user_item_train.index
train_arts1 = user_item_train.columns
match_idx = list(train_idx1.intersection(test_idx))
match_arts = list(train_arts1.intersection(test_arts))
#number of users in the test set can be predicted
print(len(match_idx))
#number of articles in the test set can be predicted
print(len(match_arts))
#number of articles in the test sets
print(len(test_arts))
#number of users in the test set
print(len(test_idx))

20
574
574
682


# Replace the values in the dictionary below
a = 662 
b = 574 
c = 20 
d = 0 


sol_4_dict = {
    'How many users can we make predictions for in the test set?': c, 
    'How many users in the test set are we not able to make predictions for because of the cold start problem?':a, 
    'How many articles can we make predictions for in the test set?': b,
    'How many articles in the test set are we not able to make predictions for because of the cold start problem?': d
}

t.sol_4_test(sol_4_dict)

Awesome job!  That's right!  All of the test articles are in the training data, but there are only 20 test users that were also in the training set.  All of the other users that are in the test set we have no data on.  Therefore, we cannot make predictions for these users using SVD.


# fit SVD on the user_item_train matrix
u_train, s_train, vt_train = np.linalg.svd(user_item_train, full_matrices=True)# fit svd similar to above then use the cells below


#find the subset of rows in the user_item_test dataset that you can predict using this matrix decomposition
train_common_idx = user_item_train.index.isin(test_idx)
train_common_col = user_item_train.columns.isin(test_arts)


u_test = u_train[train_common_idx, :]
vt_test = vt_train[:, train_common_col]
print('test: ', u_test.shape, vt_test.shape)

test:  (20, 4487) (714, 574)


# Use these cells to see how well you can use the training 
# decomposition to predict on test data


train_idx = user_item_train.index
common_idx = list(set(train_idx) & set(test_idx))
print(common_idx)

[3968, 4487, 4231, 3093, 3989, 3990, 3740, 3998, 4002, 4274, 3777, 4293, 3527, 3532, 3024, 3801, 3684, 2917, 4204, 3193]


user_item_test_subset = user_item_test.loc[common_idx]

num_latent_feats = np.arange(10,700+10,20)
train_sum_errs = []
test_sum_errs = []

for k in num_latent_feats:
    # restructure with k latent features
    s_train_new, u_train_new, vt_train_new = np.diag(s_train[:k]), u_train[:, :k], vt_train[:k, :]
    u_test_new, vt_test_new = u_test[:, :k], vt_test[:k, :]
    
    # take dot product
    user_item_train_est = np.around(np.dot(np.dot(u_train_new, s_train_new), vt_train_new))
    user_item_test_est = np.around(np.dot(np.dot(u_test_new, s_train_new), vt_test_new))
    
    # compute error for each prediction to actual value
    diffs_train = np.subtract(user_item_train, user_item_train_est)
    diffs_test = np.subtract(user_item_test_subset, user_item_test_est)
    
    # total errors and keep track of them
    train_err = np.sum(np.sum(np.abs(diffs_train)))
    train_sum_errs.append(train_err)
    
    test_err = np.sum(np.sum(np.abs(diffs_test)))
    test_sum_errs.append(test_err)


plt.plot(num_latent_feats, 1 - (np.array(test_sum_errs)/(user_item_test_subset.shape[0]*user_item_test_subset.shape[1])), label='test');
plt.legend(loc='best')
plt.xlabel('Number of Latent Features');
plt.ylabel('Accuracy');
plt.title('Accuracy vs. Number of Latent Features');
plt.show();


from subprocess import call
call(['python', '-m', 'nbconvert', 'Recommendations_with_IBM.ipynb'])

1

	article_id	title	email
0	1430.0	using pixiedust for fast, flexible, and easier...	ef5f11f77ba020cd36e1105a00ab868bbdbf7fe7
1	1314.0	healthcare python streaming application demo	083cbdfa93c8444beaa4c5f5e0f5f9198e4f9e0b
2	1429.0	use deep learning for image classification	b96a4f2e92d8572034b1e9b28f9ac673765cd074
3	1338.0	ml optimization using cognitive assistant	06485706b34a5c9bf2a0ecdac41daf7e7654ceb7
4	1276.0	deploy your python model as a restful api	f01220c46fc92c6e6b161b1849de11faacd7ccb2

	doc_body	doc_description	doc_full_name	doc_status	article_id
0	Skip navigation Sign in SearchLoading...\r\n\r...	Detect bad readings in real time using Python ...	Detect Malfunctioning IoT Sensors with Streami...	Live	0
1	No Free Hunch Navigation * kaggle.com\r\n\r\n ...	See the forest, see the trees. Here lies the c...	Communicating data science: A guide to present...	Live	1
2	☰ * Login\r\n * Sign Up\r\n\r\n * Learning Pat...	Here’s this week’s news in Data Science and Bi...	This Week in Data Science (April 18, 2017)	Live	2
3	DATALAYER: HIGH THROUGHPUT, LOW LATENCY AT SCA...	Learn how distributed DBs solve the problem of...	DataLayer Conference: Boost the performance of...	Live	3
4	Skip navigation Sign in SearchLoading...\r\n\r...	This video demonstrates the power of IBM DataS...	Analyze NY Restaurant data using Spark in DSX	Live	4

	neighbor_id	num_interactions	similarity
3869	3870	75.0	74.0
3781	3782	135.0	39.0
22	23	135.0	38.0
202	203	96.0	33.0
4458	4459	96.0	33.0
48	49	101.0	29.0
3696	3697	100.0	29.0
97	98	97.0	29.0
3763	3764	97.0	29.0
3909	3910	60.0	25.0
241	242	59.0	25.0
39	40	52.0	24.0
4931	4932	52.0	24.0
3739	3740	71.0	23.0
57	58	70.0	23.0
3595	3596	59.0	23.0
51	52	58.0	23.0
289	290	58.0	22.0
3539	3540	72.0	21.0
911	912	71.0	21.0
4784	4785	62.0	21.0
20	21	61.0	21.0
2981	2982	58.0	21.0
753	754	44.0	21.0
169	170	71.0	20.0
3168	3169	70.0	20.0
134	135	46.0	20.0
3620	3621	45.0	20.0
4641	4642	43.0	20.0
203	204	72.0	18.0

Recommendations with IBM¶

Table of Contents¶

Part I : Exploratory Data Analysis ¶

Part II: Rank-Based Recommendations ¶

Part III: User-User Based Collaborative Filtering ¶

Part IV: Content Based Recommendations (EXTRA - NOT REQUIRED)¶

This part is NOT REQUIRED to pass this project. However, you may choose to take this on as an extra way to show off your skills.¶

This part is NOT REQUIRED to pass this project. However, you may choose to take this on as an extra way to show off your skills.¶

This part is NOT REQUIRED to pass this project. However, you may choose to take this on as an extra way to show off your skills.¶

Part V: Matrix Factorization ¶

Extras¶

Conclusion¶

Directions to Submit¶

	ranked_article_id	num_ratings	title
0	1429.0	937	use deep learning for image classification
1	1330.0	927	insights from new york car accident reports
2	1431.0	671	visualize car data with brunel
3	1427.0	643	use xgboost, scikit-learn & ibm watson machine...
4	1364.0	627	predicting churn with the spss random tree alg...

article_id	0.0	2.0	4.0	8.0	9.0	12.0	14.0	15.0	16.0	18.0	...	1434.0	1435.0	1436.0	1437.0	1439.0	1440.0	1441.0	1442.0	1443.0	1444.0
user_id
1	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0
2	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3	0.0	0.0	0.0	0.0	0.0	1.0	0.0	0.0	0.0	0.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
4	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
5145	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5146	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5147	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5148	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5149	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	1.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

Recommendations with IBM¶

Table of Contents¶

Part I : Exploratory Data Analysis¶

Part II: Rank-Based Recommendations¶

Part III: User-User Based Collaborative Filtering¶

Part IV: Content Based Recommendations (EXTRA - NOT REQUIRED)¶

This part is NOT REQUIRED to pass this project. However, you may choose to take this on as an extra way to show off your skills.¶

This part is NOT REQUIRED to pass this project. However, you may choose to take this on as an extra way to show off your skills.¶

This part is NOT REQUIRED to pass this project. However, you may choose to take this on as an extra way to show off your skills.¶

Part V: Matrix Factorization¶

Extras¶

Conclusion¶

Directions to Submit¶

Part I : Exploratory Data Analysis ¶

Part II: Rank-Based Recommendations ¶

Part III: User-User Based Collaborative Filtering ¶

Part V: Matrix Factorization ¶